Alcohol dependence (AD), which is highly prevalent in the United States, is rarely treated with medications approved by the FDA. Concerted efforts to promote the use of the three FDA-approved medications for AD have had limited success, largely because of their modest efficacy. The anticonvulsant topiramate (TOP), though not approved to treat AD, substantially reduced the frequency of heavy drinking (HD) in four placebo-controlled trials and two open-label studies. Based on these findings, TOP is increasingly being prescribed off-label to treat AD (e.g., in the VA Healthcare System). Recently, we found that the ability of TOP to reduce HD was limited to individuals with the CC genotype of rs2832407, a single nucleotide polymorphism (SNP) in GRIK1, the gene encoding the kainate receptor GluK1 subunit. This finding for TOP adds to similar Pharmacogenetics findings for two other medications to treat AD, the beneficial effects of which are substantially enhanced by genetic moderators: naltrexone (which is moderated by a SNP in the mu-opioid receptor gene, OPRM1) and ondansetron (which is moderated by two genotypes in SLC6A4, the serotonin transporter gene). Consistent with the goal of personalized treatment for AD, these findings would allow clinicians to identify, in advance, which patients are likely to respond to each of these medications and which should be spared the unnecessary adverse effects that may accompany treatment in a likely non-responder. Of note, together, these three Pharmacogenetics findings would make it possible to select the best medication to reduce HD in ~75% of European Americans. METHODS: To advance the effort to develop personalized pharmacotherapy for alcohol use disorders (AUDs), we propose to conduct a 12-week, prospective, randomized clinical trial of the moderating effect of rs2832407 on the efficacy of TOP in reducing HD in 200 individuals of European descent with DSM-5 AUD. We will stratify the randomization on genotype and oversample rs2832407*C homozygote's, the most TOP-responsive genotype, to ensure comparable numbers of patients in the four medication x genotype groups. We will use daily data collection to examine changes in relevant process variables (e.g., alcohol expectancies) and their interaction with genotype and medication group as predictors of HD. The proposed study is innovative in that it will be the first prospective tes of a Pharmacogenetics hypothesis involving TOP: it will use daily reports to examine expectancies and how they interact with medication and genotype to predict HD~ and it will enroll DSM-5 AUD patients whose goal is either to reduce or stop drinking, which will increase the study's external validity. PUBLIC HEALTH IMPACT: The capacity to differentiate, in advance, likely responders from non-responders to a medication such as TOP would substantially enhance the efficacy of alcohol treatment, avoid unnecessary adverse effects, and improve the quality of care for AUD. It would also likely increase the use of a highly efficacious medication to treat AUD, which is currently undertreated and for which evidence-based treatment is underutilized.